- Email: [email protected]
Stereochemistry of addition reaction to serratene and serratenediol
Termbdmm
VoL26, pp. 751lo 761. Pagwmn
Ren
1970. Prided in &xl
Britio
STEREOCHEMISTRY OF ADDITION REACTION TO SERRATENE AND SERRATENEDIOL* Y. Tsumt
and T. SANO
Showa College of Pharmaceutical Sciences, Setagaya-ku. Tokyo, Japan.
and Y. INuB~SHI Faculty of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan. (Receiti
in Japan 9 July 1969; Received in the UK for pubiictiion I October 1969)
AWteric course of addition react&~ (hydroboration, epoxidation, osmolation, and hydrogenation) to the double bond of serratene derivatives 1 and 17 have been discussed and established as proceeding essentially from a direction. For maal reduction of the Bepoxide 3, the stereochemistry of the product is probably controlled not only by conformational stability of the intermediate carbanion but also by a steric factor of protonation to the carbanion. The Bepoxide 3 and la on treatment with acid, rearranged into new skeletal compounds (neoserratane derivatives, 26 and 27) whose stereochemistry is also diiussed. PREVIOUSLY we have reported l that serratene (l), when converted to saturated derivatives, produces two stereoisomers, a- and g-compounds, concerning the C,, In this paper, we wish to establish their stereochemistries and to clarify the steric course of addition reactions to the double bond of this particular series of compounds, and also present an example of skeletal rearrangement: 14p, 15fipoxyserratane to 15oxoneoserratane.
Steric course of the addition reactions to serratene derivatives and the stereochemistry
of the products Hydroboration and epoxidation of serratene (1) are highly stereoselective. Hydroboration of 1,’ followed by oxidation with hydrogen peroxide yielded a high proportion (-7096) of bserratan-1531 (2). Oxidation of this with Jones’ reagent gave gserratanl5one (5).’ Epoxidation of 1 with monoperphthalic acid gave a single product, 141, 15Bepoxyserratane (3). The isomeric epoxide was not found in the reaction mixture. The epoxide 3 resisted reduction with LAH but was reduced by lithium in ethylamine to afford two isomeric alcohols, m.p. 174-177O (A) and m,p. 206208O (B) in approximately equal amounts, one of which (A) was identical with @rratan15-01(2) obtained above. The alchol (B) had a secondary OH group (NMR spectrum) and was oxidized to a different ketone, a-serratanl5one (6). The two ketones could be equilibrated by alkaline treatment to give the same 4: 1 mixture of a- and bserratan15-one from either compound. The results show that the a-isomer is thermodynamically more stable than the Bisomer. Of the two possible stereoisomers (14a-H l This work was preliminarily presented at the 21st annual meeting of Pharmaceutical Society of Japan. October, 27, 1965. 1! To whom enquiries on this paper should be addressed
751
Y. TSIJDA, T. SANOand Y. ~NUBIJ~HI
CHART1
OH
7
Stereochemistry of additiin reaction to semtene and serratenediol
753
and 14pH) of U-one, the 140&l-15one should be more stable, since the compound can have all chair conformation (6a) in which the 27-methylene group is in the favoured equatorial orientation to ring D. The 14g-H isomer, however, if it had all chair conformation (Sa), would result in 1,3diaxial interaction between the 27methylene group and 28-methyl group, which could produce the conversion of ring D to the boat form (Sb). In any tax the 148-H isomer is less stable than the 14a-H isomer. It follows then that the ketone (6) has a-conliguration of hydrogen at Ci,, while 5 has 8-con8guration. Actually solvent shifts of Me groups in the NMR (THeffect) of each ketone clearly shows that the a-isomer has the 14a-H configuration with conformation 6s and the kisomer has 148-H configuration with conformation Sb: AA the alcohol (A) gave the 148H-15-one (5), this was assigned as 14@erratan15l%ol(2), since it was obtained stereoselectively by hydroboration-oxidation in which the steric course was known to be cfi-addition.4 It follows, t&refore that the con& guration of the oxide group of 3 is 8, since it gave the alcohol 2 by metal reduction. This formulation is also supported by its resistance to reduction by LAH, for a 14a, 15aepoxide should be reduced to an 14aalcohol with no special difficulty. The other alcohol (B) is therefore 14a-serratan-158-01 (4).
Reduction of epoxides with lithium and ethylamine generally proceeds by attack of solvated electrons and the structures of products obtained are determined by conformational stability of an intermediate carbanion and by rule of axial opening of epoxides2 Thus 58, 68epoxycholestane (9) affords cholestan-68-ol(l0) as a major product, while 7a, 8aepoxycholestane (11) gives a single axial product, cholestan8a01 (12), as its ring B is in a boat conformation.2 In contract to this latter example, 148, 15&epoxyserratane (3) gave no tertiary alcohol but furnished two secondary alcohols 2 and 4. If epoxide 3 has a chair conformation in ring D (this assumption was l This was preliminarily presented as the 1st Intcmational Symposium on Nuclear Magnetic Resonance (Ref. B).
754
Y. TSUDA, T. Smo
and Y.
INUJWSHI
supported by its acid rearrangement reaction+ree next section) these are the products of axial opening However, the formation of 50% of the C/D-cts isomer (2) has still to be explained, since the csrbanion will exist as an equilibration mixture 13 =14 and the proportion of 13 should be higher than that of 6 in equilibration of ketones 6 = 5 due to tetrahedral charactor of C,,. On the other hand, addition of a proton (probably solvated) to the carbanion 13 from the a-side, gaving the more stable CID-trans isomer 4, is apparently suppressed by the steric hindrance due to interaction of the 28Me group. Therefore, the overall result can be explained by balancing of two counter factors of conformational stability of an intermediate carbanion and steric hindrance of protonation to the carbanion, hence giving 2 and 4 in a ratio of approximately 1: 1. The above results show that non-catalytic addition reactions to the double bond of serratene exclusively take place from a @hrection. Reduction of a- and 8+erratan-15-one with LAH afforded two new alcohols, 8 and 7, respectively, either of which on oxidation gave back the original ketones. Again 8face attack is predominant. Osmolation of serratene (1) furnished a glycol, m.p. 255-259’, as a major product whose stereochemistry was assigened as 148, 158-dial (15) by analogy with the above evidence, presence of the minor glycol being recognized only by TLC. The same oxidation on serratenediol diacetate (17) gave an analogous result, two isomeric glycols, A (-50%) and B (-20%) being isolated in pure form after chromatography and crystallization of the product. The major glycol (A) was considered to be a 148, 15&diol(19), while the mirror one (B) to he a Ma, 15a-diol (Zo). In order to prove their configurations, they were converted to the corresponding triacetates, (21 and 22), which were subjected to Serini reaction. Generally in the Serini reaction the confignration of the C ,,-hydrogen of the resulting ketones are the same as the C ,s-hydrogen of the original triacetates.5 Reflux of each triacetate with zinc dust in toluene, however unexpectedly gave rise to the same &membered ketone 23. Sublimation of the triacetates with zinc dust gave more stereoselective results, triacetate-A (21) affording a single product (23) while triacetate-B (22) gave, as expected, a mixture of IEketones as shown by TLC. The ketone 23 should have the more stable 14a-H configuration, since this could be prepared from either glycol in excellent yield by treatment with 3% ethanolic hydrochloric acid, and under such a equilibration condition the product must be the thermcdynsmically more stable isomer. The lack of stereoselectivity in the Serini reaction may be rationalized by finding that 14~seratan15-one (5) was partly isomerized into the 14a-isomer 6 under reflex with zinc dust iu toluene containing catalytic amounts of acetic acid.
15: R = R’ = H 16: R = H, R’ = AC 19:R=OAc,R’=H 21:R=OAc,R’=Ac
2O:R=H 22: R = AC
Stercochmisby
of addition reaction to scrratenc and scrrataxdiol
755
AC
Catalytic hydrogenation of serratene (1) in acetic adid over platinum gave two stereoisomeric hydrocarbons, a- and @erratane (24 and 25), with slight excess of the latter compound.’ The stereochemistry of these hydrocarbons follows from the fact that Wolff-Kishner reduction of both 14a- and 14~serratan 15-one, (6 and 5) gave aserratane (24) as a major product.’ Under this basic condition, both ketones (6 and 5) should exist in equilibration and the resulting hydrocarbon must have the more stable C/D-trms (Ma-H) fusion. Reduction of stereoselectivity in hydrogenation is not surprising since by slow hydrogenation of a substituted double bond in a protic solvent, it has otten been observed 6*7that the stereochemistry of the product is not
24
25
controlled by ease of the approach to the catalyst surface, but hydride transfer to the protonated double bond (carbonium ion) becomes an important step thus giving the thermodynamically more stable isomer. Rearrangement of 148,15~epcxyserfatane to 15-oxoneoserratane When 14&15g-epoxyserratane (3) was treated with dilute hydrochloric acid, it was isomerixed into an aldehyde (26) together with a small amount of conjugated diene _: 231, 241, 249 and 280 mu). A singlet at S 9.50 in its NMR mixture(WL spectrum indicated the formation of an aldehyde function attached to a fully sub stituted carbon. Similarly 38,2ladiacetoxy-148, 15Bepoxyserratane (18), which was stereoselectively prepared from serratenediol diacetate (17) by epoxidation or by ozonixation, afforded the corresponding aldehyde (27; NMR signal, 6 9-55 s.) in good yield. The most probable reaction pathway from the g-oxide (3) to the aldehyde (26) is the one in which migration of C ,& ,6 bond to C,, proceeds concertedly with the breaking of the C r,-O linkage, consequently the conflguration of the aldehyde function at C ,, must be g. The mechanism via a discrete carbonium ion at C,, as pre&nted* in acid catalysed cleavage of taraxereneoxide (38) is unlikely, since if the
Y. TSUDA. T. SANO and Y. INIJBUSH~
26:R=H 21: R = OAc
%R=OH.H
B:R=O 30: R = H,
&
:
H
Rw
b/
31*R=H*R’=H
&=H*:
R’=CH
33:R=Oic,H;R’=;1 34:R=OAc,H;R’=CH, 35: R = OH, H; R’ = H 36:R=OH,H;R’=CH, 37:R=O;R’=CH,
‘.
38
carbonium ion were an intermedia% the bglycol (19) must give the same reaction product as the Bepoxide (18), whereas both the a- and ~glycols, (29 and 19), did not afford an aldehyde but gave the same 15-one (23) under the similar acidic conditions (see above). The difference in behavior of the g-oxide (18) and the jIglyco1 (19) in acidic medium was rationalized by assuming that ring D of the fSglyco1 exists in a boat conformation (39), as in 14@rratan-15one, having an anti-coplanar geometry of 14wH and 15a-H prdirred to trans elimination, and that the Bcpoxide possesses a chair conformation (49) in ring D where the migrating bond (C ,S-C J and departing group (C r,-O) have an anticoplanar geometry favorable for back side attack. The two serious non-bonded interactions between 2%Me and 27methylene groups, and between 26-Me and C ,,-OH groups in the jSglyco1 will be the chief driving forces preventing ring D from adopting chair conformation. Indeed the signal of C,,hydrogen in the NMR spectrum of pglycol(l9) appears as a broad multiplet (6 3.30) with band width 30 c/s, which indicates the hydrogen at C,, is in an axial orientation.
Stereochemistry
of addition reaction to serratene and serratenediol
151
Wolff-Kishner reduction of the aldehydediacetate (27) gave the diol(28). Oxidation of 28, followed by WoUT-Kishner reduction of the resulting diketone (29) gave the corresponding hydrocarbon (30). We now propose the name “neoserratane” for this carbon skeleton. A similar rearrangement was observed in several oxidation reactions. Oxidation of serratenediol diacetate (17) by hydrogen peroxide in formic acid directly produced the aldehyde (27), apparently the epoxide being the intermediate in this case. Oxidation of serratene (1) with chromium trioxide in acetic acid yielded neoserratan- 15-oic acid (31) as a main product (the methyl ester, m.p. 149-1 500). Transformation of the aldehyde 26 to the methyl ester 32 confirmed its structure (Experimental)
EXPERIMENTAL Unless otherwise stated, IR spectra were. taken on Nujol mull, m.ps were determined on a Yanagimoto m.p. apparatus, and NMR spectra were measured in CDCI, by a Varian A-60 machine. Identities were con6rmed by IR and by TLC comparisons. 146, l5~Epoxysewarane (3) To a soln of l(760 mg) in ether (50 ml) an ether soln of monoperphthahc acid was added. The mixture was kept in the dark for a week at room temp. Aftez being worked up as usual, the product was crystallixed from benxene-n-hexane to give 3 (400 mg) as prisms, m.p. 248-250°. (Found: C, 84.39; H, 11.86. Calc, for C,&,,,O: C. 8444; H, 11.81%). Reduction of 148, 15~egoq~rratane (3) Li (0.8 g) was added to a stirred soln of 3 (350 mg) in J?tNH, (80 ml) at 0”. Stirring was continued for a further 6 hr at O-6’. At& addition of NH,CI, the mixture was kept overnight, diluted with water, and extracttd with ether. The etheral extract was washed with water, dried over MgSO, and evaporated to give an oil which showed two spots on TLC. The separation of these was achieved by the preparative TLC [solvent: benz.en~-hexane (1:l)l. Stripping of the upper zone and extraction with CHCl, afforded 14~serrotan-15@-01 (4; 127 mg) as needles from acetone. m.p. 206207”. (Found: C. W14; H. 12.49. Calc. for C,,H,,O : C, 8404; H, 12.23%). Similarly the lower xooe gave 14~serrfztun-15gOl (2,.r 133 mg) (needles from acetone), m.p. and mixed m.p. 17&178’. Aoetylation of 4 (82 mg) with AczO (0.5 ml) and pyridine (1 ml) gave 15@~oxy-14a-serretune (needles from acetone), m.p. 186-188’. (Found: C, 81.76; H, 11.76. Calc. for C&I,,O,: C, 81.64; H, 11.56%). Similarly acetylation of 2 gave 15@zetoxy-14@erratan~’ m.p. 175-177’.
Toasolnof4(40mg)inaatone(5ml~asoln(lml)ofCrO,indilH~0,(1~03gofCrO,in30mlof water and 8.7 ml of HzSO,) was added at 0’. and the mixture was stirred for 10 mm. After decomposition of 00, by addiin of MeOH, the mixture was diluted with water, and the ppt was collected by Altrtion, then crystallixed from CHCl,-MeOH to give 6’ (30 mg) as needles, m.p. and mixed m.p. 196198’=.
(i) Compound 5 (76 mg) and 2% NaOMe-MeOH (10 ml) in henxene (5 ml) was stirred at room temp for 3 hr. The mixture was poured into water, extracted with CHCl, and the extract was evaporated to give a crystalline residue which showed two spots on a TLC. The separation was achieved by the prepaative TLC Isolvent: ben xeo+n-hexane (1: 1)l. Stripping of the upper xone and extraction with CHCI, afforded 6 (52 mg) m.p. and mixed m.p. 196198O. The lower zone gave the starting ketone 5 (10 mg) m.p. and mixed m.p. 202-204O. (ii) Compound 6 (10 mg) in 2% NaOMe-MeOH (10 ml) was kept at room temp for 4 hr. Tk product
758
Y. TSUD~ T. SANO and Y. brunosrir
isolated showed two spots on TLC corresponding to the starting ketone 6 and 5. The ratio was the same with that of the product described above.
Compound 5 (62 mg) and LAH (100 mg) in THF (10 ml) was heated under retlux for 3 hr. The oily product isolated as usual, was acetylated with AcrO (O-5 ml) and pyridine (1 ml). Crystalbxation of the product from acetone gave 15a-oceraa-y-14~serrorane (40 mg) as needles, m.p. 179-181’. (Found: C, 81.69; H, 11.26. Calc. for C&,0,: C. 8164; H, 11.56%). Treatment of this (30 mg) with LAH (40 mg) in THF (10 ml) gave 7, m.p. 95-97“, which was oxidixed with Jones’ reagent into 5 (prisms from CHCl,-MeOH), m.p. and mixed m.p. 202-204O. 14a-Serratan-15a-o1(8) Compound 6 (52 mg) was similarly reduced with LAH (100 mg) in THF (20 ml) Acetylation of the product and crystallization from acetone gave 15a-acetoxy-14x-serratane (Zs mg) as needles, m.p. 166168’. (Found: C, 81.76; H, 11.38. Calc. for C&,0,: C, 81.64; H, 11.56%). Treatment of this (15 mg) with LAH (30 mg) in THF (10 ml) gave 8, m.p. 185-190°, which was oxidized with Jones’ reagent into 6, m.p. and mixed m.p. 196198O. Osmolarion of serratene (1) A mixture of l(356 mg) and 0~0, (220 mg) in pyridine (20 ml) was kept in the dark for 5 days at room temp. H+I gas was passed into the mixture, the resulting ppt was filtered and washed with CHCI, The combined filtrate and washings were evaporated to give a crystalline residue which was chromatographed in benxene-n-hexane (1: 1) over AlxO, (4 g). First elution with the same solvent (10 ml) gave an oil (30 mg) which was discarded Furtba elution with the BBM solvent (50 mI) gave a solid (62 ma) which showed two spots on TLC. The lower major spot corresponded to 15 described below. Elution with benxene gave a crystalline residue (179 mg) which was crystallixed from CHCI,MeOH to give 148,lSg dihydrqwrmtane (15; 120 mg) as needles, m.p. 255-259’. Acetylation of this (80 mg) with AcrO (O-5 ml) and pyridine (1 ml) gave 16 (needles from CHCI,MeOH), m.p. 217-220’. Osmolotionof sewotenediol diacetote (17) Compound 17 (2.2 g) and OsO, (1.0 g) in pyridine (120 ml) was kept in the dark for a month. An
aqueoussoln(9Oml)ofNaHSO,(5~6g)wasthenaddedandtbemixturewasstirredfor30minatroom temp, and extracted with CHCI, The extract was washed with water and evaporated to dryness br vocuo to give a crystalline residue (2.3 g). which was washed with benxene and the insoluble part (1.2 g) was crystallized from CHCl,-MeOH to give 38, 2la-&ceraryl4g, lS~Ury&o~yserrororte (19; 0.8 g) as needles, m.p. 316318’ (open capillary). NMR (6): 0.87 (15H. 8). 1.00 (3H, s), l-19 (3H. s), 2.10 (6H, s), 3.70(1H, m), 4.50(2H,m).(Found: C, 72.32;H,9.75. Calc.forC$&,:C, 72.82;H, 10.06%). The henxene soluble part was chromatographed in benxene over alumina and separated as follows, (i) Elution with benzene (100 ml) gave the starting material (0.4 g). (ii) Further elution with benxene (1 1.) gave 38,2ladiaceroxy-14% 15adfJrydrrqserrotrme (20; 0.54 g) (needles from benxene-n-hexane), m.p. 296298’ (open capillary); NMR (6): 0.89 (9H. s), O-92 (6H, s), 099 (3H. s), l-01 (3H. s). 2.10 (6H, s), 3-30 (IH, m), 4-50 (lH, m). (Found: C, 72.71; H, 10.30. Calc. for Cr,H,p,: C, 72.82; H, 10.06%). (iii) Further elution with CHCI, (300 ml) gave 19 (0.4 g) as needles, m.p. and mixed m.p. 316-318O (open caphary). Acetylation of 19 (100 mg) with Ac,O (1 ml) and pyridine (2 ml) gave 36. 156, 2la-rrloceroxy-14b hydroaysewatane (21; 95 Ing) as needles from benzene-n-hexane, m-p. 297-299’. (Found: C, 71.68; H, 9.63. Calc. for Ca,,O,: C, 71.72; H, 9.70%). Acetylation of 20 (100 mg) with AcsO (1 ml) and pyridine (2 ml) gave 22 (97 me) as needles from bcnxene-n-hexanc, m.p. 257-259“. (Found: C, 71.64; H, 9.77. Clac. for C,$I&,: C, 71.72; H, 9.70%). Serini reaction of 16
Compound 16 (12 mg) and Zn dust (242 mg) were mixed and sublimed at 250’ and 1 mmHg. The TLV and IR of the crystalline sublimate showed that the product was a mixture of Ma and 14@-serratanM-ones, the major product being 6
Stereochemistry of addition reaction to serratene and serratenediol
759
Serini reaction of 21 and 22 (i) Compound 21 (60 mg) and Zn dust (1 g) in toluene (10 ml) were heated under retlux for 15 hr. removal of Zn dust and solvent letI a crystalline residue which was crystallized from CHCls-MeOH to give 38,2ladiac&q-l4a-~watcmtl5-one (23; 30 mg) as prisms, m.p. 296296’ (open capillary); IR 1725 @AC) and 1960 cm-’ (ketone); NMR (6): 0.85 (12H, s), 0.90 (3H. s). 0.92 (3H, s), 1.06 (3H, s), 2.05 (6iH, s). 4.45 (W, m). (Found: C, 75.52; H. 10.13. Calc. for C&,0,: C, 75.23; H, 10.03%). (ii) Compound 22 (60 mg) and Zn dust (1 g) in toluene (10 ml) were similarly treated Crystallization of the product from CHCI,-MeCH afforded 23 (20 mg) m.p. and mixed m.p. 296296”. (iii) A mixture of 21(10 mg) and Zn dust (120 mg) was sublimed at 270° and 1 mm. The TLC of the crystalline sublimate showed a single spot corresponding to 23. The IR spectrum was superimposable on that of the sample obtahred above. (iv)Amixhrreof22(11mg)andZndust(17lmg)wassublimedasdescribedabove.The~Cofthe sublimate showed two spots, one of which was identical with that of 23. The IR spwtntm of the mixture exhibited a CO band at 1685 cm-*, but no OH band. Isomcrixutfon of 14$-serratmr-l5-one under the condition #the Serid reaction Compound 5 (10 mg) and Zn dust (500 mg) in toluene (10 ml) containing 2 drops of AcOH. was re5uxed for 5 hr. The TLC of the product showed two spots corresponding to 6 and 5. 38,2 la-Diuceto~-l4a-smatan-l5-one (23) (i)Asolnof19(17mg)in3%HCCEtOH(lOml)washeatcdundarcfluxforlhr.Themixturcwas neutralized with 5% KxCO,aq and extracted with CHCI,. The extract was washed and water, dried over MgSO, and evaporated to dryness. The residue was acetylated with AcrO (1 ml) and pyridine (2 ml) at room temp to give 23 (11 mg), m.p. and mixed m.p. 294-296’ (open capillary). (ii) A sohr of 22 (40 mg) in 3 % HCI-EtOH (20 ml) and CHCI, (2 ml) was treated as described above. Acetylation of the product gave 23 (26 mg). m.p. 294-296’ (open capillary). 38, 21~Dhcetary-14B, 15~qrserrctune (18) (i) Coupound 17’ (2.4 g) in CHCI, (30 ml) was treated with an ethereal sohr of monoperphthahc acid (large excess) at room temp for a week. Tbe reaction mixture was dilti with CHCI, and the organic layer was washed with 5% NaOH aq, water, dried over MgSO, and evaporated to dryness. Crystallixation of the residue from benzene afforded the oxide (18) as prisms (1.58 g). m.p. 350-353” (open capifbuy), [a]n +27.8” (c = 36 in CHCI,); NMR (a): 079 (3H s), 090 (12H s), 097 (3H, s), 1W (3H, s), 2.10 (6H, s), 2.88 (broad s), 450 (ZH, m) (Found: C, 7508; H, 9.83. Calc for Cs,H,,O, : C, 75.23; H, 10.03%). (ii) Oxonixed 0, was passed through a soln of 17 (728 mg) in CHCI, (15 ml) at 0’ until the mixture was s&mated to tetranitromethane. The mixture was diluted with water, extramd with CHCI, and the extract was washed with water, dried over MgSO, and evaporated to dryness. Crystallization of the residue from CHCI,-MeCH gave 18 (600 mg) as prisms, m.p. and mixed m.p. 340-343”. 15-oxoneoserrataIu? (26) Compound 3 (60 mg) in CHCl, (I5 ml) was treated with a few drops of cone HCl under stirring for 30 min. and the product in n-hexane was chromatographed over alumina (1 g). The Erst ehmte afforded a hydrocarbon mixture (10 mg), 1, (EtCH): 23 1,241,249 and 280 mp, further elution with n-hexane and benzene gave the aldehyde (W, 35 mg) as prisms from acetone, m.p. 17617S”; IR: 2703 and 1718 cm-’ (CHO); NMR (a): 0.70 (3H, s), 0.76 (9H. s), 0.79 (12H. s), 9.50 (lH, s). (Found: C, 84.68; H, 11.98. Calc. for C,&,O: C, 8444; H, 11.81%). 38,2la-Diacet~l5-oxoneoserrctane (27) (i)Compounds 18(198mg)inCHC1,(15ml)wastreatedwithconcHC1(0~1mf)atroomtempfor30 min with stirring. Crystallization of the product kom benxene atTorded the uldehyak (27; 90 mg) as needles, m.p. 245-247”; [a], +21.6” (c = 3.43 in CHClJ; IR: 1730 cm-* (CHO, OAc); NMR (a): 0.75 (3H, s), 0.82 (3H, s), 0.88 (9H, s), 0.96 (6H. s), 2.09 (6H. s), 450 (W. m). 9.55 (lH, s). (Found: C, 75.14: H, 9.84. Calc. for CjI,,O,: C, 75.23; H, 10.03%). (ii)Toasolnof17(500mg)inAcOEt(5Oml)wasaddcdasdnof8O%HCOOH(l0ml)and3O% H,O,42ml),andthemixtunwasheatedundcrreeuxfor lhr,and concenkatedtoca.+vohuneinvatxW.
760
Y. TSUDA,T. SANOand Y.
hUBUSHI
Collection of the ppt and crystallization from CHCI,MeOH gave 27 (260 mg), as leaflets, m.p. and mixed m.p. 24s246O. (iii) Compound 17 (728 mg) in CHCI, (15 ml) was ozonized as described above. Zn dust (1.1 g) and AcOH (8 ml) were added to the mixture, and after. stirring for 2 hr the ppt was Eltered off and washed with CHCI, The combined filtrate and washings were washed with water, dried over MgSO, and evaporated to dryness. Chromatography of the product and crystallization from CHCl,MeOH gave 27 (350 mg), m.p. and mixed m.p. 247-248”. Hydrolysis of 27 (100 mg) with 2% KOH-MeOH (30 ml) under refluz for 1 hr afforded the corresponding diol (70 mg) as needles from MeOH m.p. 209-211°; IR: 3390 cm-* (OH), 1721 (CHO). (Found : C, 7550; H, 10.89. Calc. for C,,,H,,O, . HsO : C, 75.58 ; H, 11GCl %). 3 B, 2 la-Diaceto~eoserratan- 15-oic acid (33) To a soln of 27 (290 mg) in pyridine (10 ml) was added KMnO, (240 mg) and the mixture was stirred for 7 hr at room temp. After decomposition of KMnO, by MeOH, the solvent was evaporated to dryness in D(UIUO. The residue was taken up in CHQ, and the extract was washed with Na&O, and dil HsSO,, water, dried over MgSO,, and evaporated to dryness Crystallization of the residue from MeOH afforded the crcid (33; 170 mg) as needles, m.p. 285-2XP, [o]u -5.1” (c = 2.7 in CHCl,); IR: 32541705 (COOH), and 1726 cm-’ (OAc). (Found: C, 72.95; H, 9.85. Calc. for CJi,,O,: C, 73.08; H, 9.74%). Tbe acid 33 (110 mg) was methylated with excess of CHsNN,in ether. Crystallization of the product from CHCl,-MeOH gave the methyl ester (34; 85 mg) as needles, m.p. 232-234’, [a], -17. lo (c = 2.8 in CHCls); IR: 1730 cm-’ (COOMe, OAc). (Found: C, 7328; H, 1011. Calc for CssH,eOs: C, 73.39; H, 9.85%). 3 b 2 la-Dihydrogmeoserratane-15-o& acid (35) The diacetate 33 (115 mg) was hydrolysed with 10% KOH-MeOH (15 ml) under reiluz for 4 hr. The mixture was poured into water, and crystallization of the pp1 from CHCl,MeOH gave the dti (36; 85 mg) as needles, m.p. 345-347O (open capillary). (Found: C, 73.42; H, 10.72. Calc. for CjI,O,.H,O: C, 73.12; H, 10.64%). The acid 35 (320 mg) in CHCI,-MeOH (1:l) (30 ml) was methylated with CHsNN,in ether. Crystallization of the product from CHCI,-MeOH gave the methyl ester (36; 292 mg) as needles, m.p. 255257’; IR: 3390, 3322 (OH) and 1718 cm-’ (eater). (Found: C, 74.49; H, 10.83. Calc, for CJ-&,O,.+ H,O: C, 74.79; H, 10.73%). The diketoesfer (37) A mixture of 36 (200 mg) and 00, (400 mg) in pyridine (5 m,) was kept overnight at room temp. The product, isolated with CHCI, was chromatographed in benzene over alumina (2 g), and the eluate was crystallized from n-hexane to give the diketo-ester(37) as needles, m.p. 203-205O, [al, +35.6” (c = 3.6 in CHCI,); IR: 1709 cm-’ (ester and ketone). (Found: C, 7658; H, 991. Caic for Cs,H,O4: C 7681; H, 9.98 %). 3 g, 2 la-Dihydroxyneoserratwuz(28) Compound 27 (500 mg) and anhyd hydraxine (4 ml) in diethykneglycol(5O ml) containing Na (4 g) were heated at 180” for 4 hr, then the temp was raised to 210” and the mixture was kept for 4 hr. The mixture was poured into water, extracted with CHCl, and the extract was washed with water, dried over MgSO, and evaporated to dryness. Crystallixation of the residue (420 mg) from CHCI,-MeOH gave 28 as prisms, m.p. 265-267O. (Found: C. 79.07; H, 11.52. Calc. for CJI,,O,+H,O: C, 79.40; H, 11.78%). Acetylation of this (100 mg) with AC,0 (1 ml) and pyridirc (2 ml) gave the diucetute (92 mg) as needles from CHCl,MeOH, m.p. 256257”. (Found: C, 77.40; H, 10.88. Calc. for CJi,p,: C, 77.22; H, 10.67%). Neoserrotaw (30)
Compound 28 (200 mg) and 00s (300 mg) in pyridine (5 ml) were stirred overnight at room temp. The mixture was poured into water and extracted with C!HC&.The extract was washed with water, dried over Mg!SO*,and evaporated to dryrxsa The residue was chromatographal in CHCls over silica-gel to
Stereochemistry of addition reaction to serratetie and serratenediol
761
give neoserrarun-3,2l-&ne (299; 160 mg) (needles from CHCl,-MeOH), m.p. 140-143”. IR: 1695 cm-’ cyclobexanone). Compound 29 (80 mg) and anhyd hydrazinc (1 ml) in diethyleneglycol(20 ml) containing Na (1 g) were beated as described above. Chromatography of the product in n-hexane over A1203 gave 30 (40 mg) (leaflus from CHCl,-M&H). (Found: C, 87.08; H. 12.56. Calc. for C&I,,: C, 87.30; H, 12.70%). Methyl neawrratan- 1S-oate (32) (i) The diketo-&er 37 (623 mg) and anhyd hydraxine (6 ml) in diethyleneglycol(42 ml) containing Na (3 g) were heated at MO0 for 4 hr, and heating was continued at 210” for further 2 hr. The mixture was poured into water and extracted with CHCI, The extract was washed with water, dried over MgSO, and evaporated to dryness. The residue in ether (10 ml) was treated with CHxNN,to give the methyl ester (32; 510 mg) as needles from MeOH, m.p. 149-151°; IR: 3509 (OH) and 1724 cm-’ (COOMe). (Found: C, 79.89; H, 11.49. Calc. for C,,H,xO,.+MeOH: 80.03; H, 11.50%). (ii) The aldehyde 26 (100 mg) was oxidized with KMnO, in pyridine as described for 33. Methylation of the crude product with CHgN, afforded the methyl ester 32, m.p. and mixed m.p. 149-151°. Oxidation of serratene (1) with chromium trioxiak Smatene 1(467 mg) in AcOH (150 ml) and 00, (500 mg) in 90% AcOH (10 ml) were mixed and the mixture was warmed at 75O on a water bath for 3 hr. At&r decomposition of excess of 00, with EtOH, AcOH was removed &I vucno, and the residue was taken up in CHCI, which was washed with water, dried over MgSO, and evaporated to dryness. Washing of the residue with n-hexane gave a crystalline residue (225 mg). Chromatography of this and crystallixation of benzene and CHCI, eluates gave neosewatnn-15-o& acti (31; 210 mg) (Rims from CHCl,-MeOH), m.p. >300°; IR: 1690 cm-’ (COOH). (Found: C, 81.42; H, 11.34. Calc. for CjI,O,: C, 81.39; H, 11.38%). Treatment of 31(50 mg) with CHflN, in ether gave 32 (42 mg) as needles from CHCI,-MeOH, m.p. and mixed m.p. 149-151’. The hexane-soluble oil (285 mg) was passed through a column of alumina and the eluate was crystallized from acetone to give serrat-13-en-15one’ as leaflets, m.p. and mixed m.p. 245-246’. REFERENCES i Y. Inubushi, T. Tsuda, T. Sane, T. Konita, S. Suzuki, H. Ageta and Y. Otake, Chem. Phunn. Bull. Tdyo 15, 1153 (1967). * A. S. Hallsworth and H. B. Henbest, 1. Chem. Sot. 4604 (1957). a M. Hashimoto and Y. Tsuda, Reiiminmy Report of NMR Symposium, M-2-13, Tokyo, September 1 (1965). ’ Organic Reactions Vol. 13, p 1, Wiley, N.Y. (1963). ’ L. F. Fieser and M. Fieser. S:ervids p 628. Reinhold N.Y. (1959). 6 J. W. ApSimon, P. V. Demarco and I. Len&e, Camad. J. Chin, 43.2793 (1965). ’ Y. Tsuda, K. Isobe, S. Fukushima, H. Ageta and K. Iwata, Terrohedron Letters 23 (1967). * J. M. Beaton, F. S. Spring, R. Stevenson and J. L. Stewart,I. Chem. Sot. 2131 (1955).
VoL26, pp. 751lo 761. Pagwmn
Ren
1970. Prided in &xl
Britio
STEREOCHEMISTRY OF ADDITION REACTION TO SERRATENE AND SERRATENEDIOL* Y. Tsumt
and T. SANO
Showa College of Pharmaceutical Sciences, Setagaya-ku. Tokyo, Japan.
and Y. INuB~SHI Faculty of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan. (Receiti
in Japan 9 July 1969; Received in the UK for pubiictiion I October 1969)
AWteric course of addition react&~ (hydroboration, epoxidation, osmolation, and hydrogenation) to the double bond of serratene derivatives 1 and 17 have been discussed and established as proceeding essentially from a direction. For maal reduction of the Bepoxide 3, the stereochemistry of the product is probably controlled not only by conformational stability of the intermediate carbanion but also by a steric factor of protonation to the carbanion. The Bepoxide 3 and la on treatment with acid, rearranged into new skeletal compounds (neoserratane derivatives, 26 and 27) whose stereochemistry is also diiussed. PREVIOUSLY we have reported l that serratene (l), when converted to saturated derivatives, produces two stereoisomers, a- and g-compounds, concerning the C,, In this paper, we wish to establish their stereochemistries and to clarify the steric course of addition reactions to the double bond of this particular series of compounds, and also present an example of skeletal rearrangement: 14p, 15fipoxyserratane to 15oxoneoserratane.
Steric course of the addition reactions to serratene derivatives and the stereochemistry
of the products Hydroboration and epoxidation of serratene (1) are highly stereoselective. Hydroboration of 1,’ followed by oxidation with hydrogen peroxide yielded a high proportion (-7096) of bserratan-1531 (2). Oxidation of this with Jones’ reagent gave gserratanl5one (5).’ Epoxidation of 1 with monoperphthalic acid gave a single product, 141, 15Bepoxyserratane (3). The isomeric epoxide was not found in the reaction mixture. The epoxide 3 resisted reduction with LAH but was reduced by lithium in ethylamine to afford two isomeric alcohols, m.p. 174-177O (A) and m,p. 206208O (B) in approximately equal amounts, one of which (A) was identical with @rratan15-01(2) obtained above. The alchol (B) had a secondary OH group (NMR spectrum) and was oxidized to a different ketone, a-serratanl5one (6). The two ketones could be equilibrated by alkaline treatment to give the same 4: 1 mixture of a- and bserratan15-one from either compound. The results show that the a-isomer is thermodynamically more stable than the Bisomer. Of the two possible stereoisomers (14a-H l This work was preliminarily presented at the 21st annual meeting of Pharmaceutical Society of Japan. October, 27, 1965. 1! To whom enquiries on this paper should be addressed
751
Y. TSIJDA, T. SANOand Y. ~NUBIJ~HI
CHART1
OH
7
Stereochemistry of additiin reaction to semtene and serratenediol
753
and 14pH) of U-one, the 140&l-15one should be more stable, since the compound can have all chair conformation (6a) in which the 27-methylene group is in the favoured equatorial orientation to ring D. The 14g-H isomer, however, if it had all chair conformation (Sa), would result in 1,3diaxial interaction between the 27methylene group and 28-methyl group, which could produce the conversion of ring D to the boat form (Sb). In any tax the 148-H isomer is less stable than the 14a-H isomer. It follows then that the ketone (6) has a-conliguration of hydrogen at Ci,, while 5 has 8-con8guration. Actually solvent shifts of Me groups in the NMR (THeffect) of each ketone clearly shows that the a-isomer has the 14a-H configuration with conformation 6s and the kisomer has 148-H configuration with conformation Sb: AA the alcohol (A) gave the 148H-15-one (5), this was assigned as 14@erratan15l%ol(2), since it was obtained stereoselectively by hydroboration-oxidation in which the steric course was known to be cfi-addition.4 It follows, t&refore that the con& guration of the oxide group of 3 is 8, since it gave the alcohol 2 by metal reduction. This formulation is also supported by its resistance to reduction by LAH, for a 14a, 15aepoxide should be reduced to an 14aalcohol with no special difficulty. The other alcohol (B) is therefore 14a-serratan-158-01 (4).
Reduction of epoxides with lithium and ethylamine generally proceeds by attack of solvated electrons and the structures of products obtained are determined by conformational stability of an intermediate carbanion and by rule of axial opening of epoxides2 Thus 58, 68epoxycholestane (9) affords cholestan-68-ol(l0) as a major product, while 7a, 8aepoxycholestane (11) gives a single axial product, cholestan8a01 (12), as its ring B is in a boat conformation.2 In contract to this latter example, 148, 15&epoxyserratane (3) gave no tertiary alcohol but furnished two secondary alcohols 2 and 4. If epoxide 3 has a chair conformation in ring D (this assumption was l This was preliminarily presented as the 1st Intcmational Symposium on Nuclear Magnetic Resonance (Ref. B).
754
Y. TSUDA, T. Smo
and Y.
INUJWSHI
supported by its acid rearrangement reaction+ree next section) these are the products of axial opening However, the formation of 50% of the C/D-cts isomer (2) has still to be explained, since the csrbanion will exist as an equilibration mixture 13 =14 and the proportion of 13 should be higher than that of 6 in equilibration of ketones 6 = 5 due to tetrahedral charactor of C,,. On the other hand, addition of a proton (probably solvated) to the carbanion 13 from the a-side, gaving the more stable CID-trans isomer 4, is apparently suppressed by the steric hindrance due to interaction of the 28Me group. Therefore, the overall result can be explained by balancing of two counter factors of conformational stability of an intermediate carbanion and steric hindrance of protonation to the carbanion, hence giving 2 and 4 in a ratio of approximately 1: 1. The above results show that non-catalytic addition reactions to the double bond of serratene exclusively take place from a @hrection. Reduction of a- and 8+erratan-15-one with LAH afforded two new alcohols, 8 and 7, respectively, either of which on oxidation gave back the original ketones. Again 8face attack is predominant. Osmolation of serratene (1) furnished a glycol, m.p. 255-259’, as a major product whose stereochemistry was assigened as 148, 158-dial (15) by analogy with the above evidence, presence of the minor glycol being recognized only by TLC. The same oxidation on serratenediol diacetate (17) gave an analogous result, two isomeric glycols, A (-50%) and B (-20%) being isolated in pure form after chromatography and crystallization of the product. The major glycol (A) was considered to be a 148, 15&diol(19), while the mirror one (B) to he a Ma, 15a-diol (Zo). In order to prove their configurations, they were converted to the corresponding triacetates, (21 and 22), which were subjected to Serini reaction. Generally in the Serini reaction the confignration of the C ,,-hydrogen of the resulting ketones are the same as the C ,s-hydrogen of the original triacetates.5 Reflux of each triacetate with zinc dust in toluene, however unexpectedly gave rise to the same &membered ketone 23. Sublimation of the triacetates with zinc dust gave more stereoselective results, triacetate-A (21) affording a single product (23) while triacetate-B (22) gave, as expected, a mixture of IEketones as shown by TLC. The ketone 23 should have the more stable 14a-H configuration, since this could be prepared from either glycol in excellent yield by treatment with 3% ethanolic hydrochloric acid, and under such a equilibration condition the product must be the thermcdynsmically more stable isomer. The lack of stereoselectivity in the Serini reaction may be rationalized by finding that 14~seratan15-one (5) was partly isomerized into the 14a-isomer 6 under reflex with zinc dust iu toluene containing catalytic amounts of acetic acid.
15: R = R’ = H 16: R = H, R’ = AC 19:R=OAc,R’=H 21:R=OAc,R’=Ac
2O:R=H 22: R = AC
Stercochmisby
of addition reaction to scrratenc and scrrataxdiol
755
AC
Catalytic hydrogenation of serratene (1) in acetic adid over platinum gave two stereoisomeric hydrocarbons, a- and @erratane (24 and 25), with slight excess of the latter compound.’ The stereochemistry of these hydrocarbons follows from the fact that Wolff-Kishner reduction of both 14a- and 14~serratan 15-one, (6 and 5) gave aserratane (24) as a major product.’ Under this basic condition, both ketones (6 and 5) should exist in equilibration and the resulting hydrocarbon must have the more stable C/D-trms (Ma-H) fusion. Reduction of stereoselectivity in hydrogenation is not surprising since by slow hydrogenation of a substituted double bond in a protic solvent, it has otten been observed 6*7that the stereochemistry of the product is not
24
25
controlled by ease of the approach to the catalyst surface, but hydride transfer to the protonated double bond (carbonium ion) becomes an important step thus giving the thermodynamically more stable isomer. Rearrangement of 148,15~epcxyserfatane to 15-oxoneoserratane When 14&15g-epoxyserratane (3) was treated with dilute hydrochloric acid, it was isomerixed into an aldehyde (26) together with a small amount of conjugated diene _: 231, 241, 249 and 280 mu). A singlet at S 9.50 in its NMR mixture(WL spectrum indicated the formation of an aldehyde function attached to a fully sub stituted carbon. Similarly 38,2ladiacetoxy-148, 15Bepoxyserratane (18), which was stereoselectively prepared from serratenediol diacetate (17) by epoxidation or by ozonixation, afforded the corresponding aldehyde (27; NMR signal, 6 9-55 s.) in good yield. The most probable reaction pathway from the g-oxide (3) to the aldehyde (26) is the one in which migration of C ,& ,6 bond to C,, proceeds concertedly with the breaking of the C r,-O linkage, consequently the conflguration of the aldehyde function at C ,, must be g. The mechanism via a discrete carbonium ion at C,, as pre&nted* in acid catalysed cleavage of taraxereneoxide (38) is unlikely, since if the
Y. TSUDA. T. SANO and Y. INIJBUSH~
26:R=H 21: R = OAc
%R=OH.H
B:R=O 30: R = H,
&
:
H
Rw
b/
31*R=H*R’=H
&=H*:
R’=CH
33:R=Oic,H;R’=;1 34:R=OAc,H;R’=CH, 35: R = OH, H; R’ = H 36:R=OH,H;R’=CH, 37:R=O;R’=CH,
‘.
38
carbonium ion were an intermedia% the bglycol (19) must give the same reaction product as the Bepoxide (18), whereas both the a- and ~glycols, (29 and 19), did not afford an aldehyde but gave the same 15-one (23) under the similar acidic conditions (see above). The difference in behavior of the g-oxide (18) and the jIglyco1 (19) in acidic medium was rationalized by assuming that ring D of the fSglyco1 exists in a boat conformation (39), as in 14@rratan-15one, having an anti-coplanar geometry of 14wH and 15a-H prdirred to trans elimination, and that the Bcpoxide possesses a chair conformation (49) in ring D where the migrating bond (C ,S-C J and departing group (C r,-O) have an anticoplanar geometry favorable for back side attack. The two serious non-bonded interactions between 2%Me and 27methylene groups, and between 26-Me and C ,,-OH groups in the jSglyco1 will be the chief driving forces preventing ring D from adopting chair conformation. Indeed the signal of C,,hydrogen in the NMR spectrum of pglycol(l9) appears as a broad multiplet (6 3.30) with band width 30 c/s, which indicates the hydrogen at C,, is in an axial orientation.
Stereochemistry
of addition reaction to serratene and serratenediol
151
Wolff-Kishner reduction of the aldehydediacetate (27) gave the diol(28). Oxidation of 28, followed by WoUT-Kishner reduction of the resulting diketone (29) gave the corresponding hydrocarbon (30). We now propose the name “neoserratane” for this carbon skeleton. A similar rearrangement was observed in several oxidation reactions. Oxidation of serratenediol diacetate (17) by hydrogen peroxide in formic acid directly produced the aldehyde (27), apparently the epoxide being the intermediate in this case. Oxidation of serratene (1) with chromium trioxide in acetic acid yielded neoserratan- 15-oic acid (31) as a main product (the methyl ester, m.p. 149-1 500). Transformation of the aldehyde 26 to the methyl ester 32 confirmed its structure (Experimental)
EXPERIMENTAL Unless otherwise stated, IR spectra were. taken on Nujol mull, m.ps were determined on a Yanagimoto m.p. apparatus, and NMR spectra were measured in CDCI, by a Varian A-60 machine. Identities were con6rmed by IR and by TLC comparisons. 146, l5~Epoxysewarane (3) To a soln of l(760 mg) in ether (50 ml) an ether soln of monoperphthahc acid was added. The mixture was kept in the dark for a week at room temp. Aftez being worked up as usual, the product was crystallixed from benxene-n-hexane to give 3 (400 mg) as prisms, m.p. 248-250°. (Found: C, 84.39; H, 11.86. Calc, for C,&,,,O: C. 8444; H, 11.81%). Reduction of 148, 15~egoq~rratane (3) Li (0.8 g) was added to a stirred soln of 3 (350 mg) in J?tNH, (80 ml) at 0”. Stirring was continued for a further 6 hr at O-6’. At& addition of NH,CI, the mixture was kept overnight, diluted with water, and extracttd with ether. The etheral extract was washed with water, dried over MgSO, and evaporated to give an oil which showed two spots on TLC. The separation of these was achieved by the preparative TLC [solvent: benz.en~-hexane (1:l)l. Stripping of the upper zone and extraction with CHCl, afforded 14~serrotan-15@-01 (4; 127 mg) as needles from acetone. m.p. 206207”. (Found: C. W14; H. 12.49. Calc. for C,,H,,O : C, 8404; H, 12.23%). Similarly the lower xooe gave 14~serrfztun-15gOl (2,.r 133 mg) (needles from acetone), m.p. and mixed m.p. 17&178’. Aoetylation of 4 (82 mg) with AczO (0.5 ml) and pyridine (1 ml) gave 15@~oxy-14a-serretune (needles from acetone), m.p. 186-188’. (Found: C, 81.76; H, 11.76. Calc. for C&I,,O,: C, 81.64; H, 11.56%). Similarly acetylation of 2 gave 15@zetoxy-14@erratan~’ m.p. 175-177’.
Toasolnof4(40mg)inaatone(5ml~asoln(lml)ofCrO,indilH~0,(1~03gofCrO,in30mlof water and 8.7 ml of HzSO,) was added at 0’. and the mixture was stirred for 10 mm. After decomposition of 00, by addiin of MeOH, the mixture was diluted with water, and the ppt was collected by Altrtion, then crystallixed from CHCl,-MeOH to give 6’ (30 mg) as needles, m.p. and mixed m.p. 196198’=.
(i) Compound 5 (76 mg) and 2% NaOMe-MeOH (10 ml) in henxene (5 ml) was stirred at room temp for 3 hr. The mixture was poured into water, extracted with CHCl, and the extract was evaporated to give a crystalline residue which showed two spots on a TLC. The separation was achieved by the prepaative TLC Isolvent: ben xeo+n-hexane (1: 1)l. Stripping of the upper xone and extraction with CHCI, afforded 6 (52 mg) m.p. and mixed m.p. 196198O. The lower zone gave the starting ketone 5 (10 mg) m.p. and mixed m.p. 202-204O. (ii) Compound 6 (10 mg) in 2% NaOMe-MeOH (10 ml) was kept at room temp for 4 hr. Tk product
758
Y. TSUD~ T. SANO and Y. brunosrir
isolated showed two spots on TLC corresponding to the starting ketone 6 and 5. The ratio was the same with that of the product described above.
Compound 5 (62 mg) and LAH (100 mg) in THF (10 ml) was heated under retlux for 3 hr. The oily product isolated as usual, was acetylated with AcrO (O-5 ml) and pyridine (1 ml). Crystalbxation of the product from acetone gave 15a-oceraa-y-14~serrorane (40 mg) as needles, m.p. 179-181’. (Found: C, 81.69; H, 11.26. Calc. for C&,0,: C. 8164; H, 11.56%). Treatment of this (30 mg) with LAH (40 mg) in THF (10 ml) gave 7, m.p. 95-97“, which was oxidixed with Jones’ reagent into 5 (prisms from CHCl,-MeOH), m.p. and mixed m.p. 202-204O. 14a-Serratan-15a-o1(8) Compound 6 (52 mg) was similarly reduced with LAH (100 mg) in THF (20 ml) Acetylation of the product and crystallization from acetone gave 15a-acetoxy-14x-serratane (Zs mg) as needles, m.p. 166168’. (Found: C, 81.76; H, 11.38. Calc. for C&,0,: C, 81.64; H, 11.56%). Treatment of this (15 mg) with LAH (30 mg) in THF (10 ml) gave 8, m.p. 185-190°, which was oxidized with Jones’ reagent into 6, m.p. and mixed m.p. 196198O. Osmolarion of serratene (1) A mixture of l(356 mg) and 0~0, (220 mg) in pyridine (20 ml) was kept in the dark for 5 days at room temp. H+I gas was passed into the mixture, the resulting ppt was filtered and washed with CHCI, The combined filtrate and washings were evaporated to give a crystalline residue which was chromatographed in benxene-n-hexane (1: 1) over AlxO, (4 g). First elution with the same solvent (10 ml) gave an oil (30 mg) which was discarded Furtba elution with the BBM solvent (50 mI) gave a solid (62 ma) which showed two spots on TLC. The lower major spot corresponded to 15 described below. Elution with benxene gave a crystalline residue (179 mg) which was crystallixed from CHCI,MeOH to give 148,lSg dihydrqwrmtane (15; 120 mg) as needles, m.p. 255-259’. Acetylation of this (80 mg) with AcrO (O-5 ml) and pyridine (1 ml) gave 16 (needles from CHCI,MeOH), m.p. 217-220’. Osmolotionof sewotenediol diacetote (17) Compound 17 (2.2 g) and OsO, (1.0 g) in pyridine (120 ml) was kept in the dark for a month. An
aqueoussoln(9Oml)ofNaHSO,(5~6g)wasthenaddedandtbemixturewasstirredfor30minatroom temp, and extracted with CHCI, The extract was washed with water and evaporated to dryness br vocuo to give a crystalline residue (2.3 g). which was washed with benxene and the insoluble part (1.2 g) was crystallized from CHCl,-MeOH to give 38, 2la-&ceraryl4g, lS~Ury&o~yserrororte (19; 0.8 g) as needles, m.p. 316318’ (open capillary). NMR (6): 0.87 (15H. 8). 1.00 (3H, s), l-19 (3H. s), 2.10 (6H, s), 3.70(1H, m), 4.50(2H,m).(Found: C, 72.32;H,9.75. Calc.forC$&,:C, 72.82;H, 10.06%). The henxene soluble part was chromatographed in benxene over alumina and separated as follows, (i) Elution with benzene (100 ml) gave the starting material (0.4 g). (ii) Further elution with benxene (1 1.) gave 38,2ladiaceroxy-14% 15adfJrydrrqserrotrme (20; 0.54 g) (needles from benxene-n-hexane), m.p. 296298’ (open capillary); NMR (6): 0.89 (9H. s), O-92 (6H, s), 099 (3H. s), l-01 (3H. s). 2.10 (6H, s), 3-30 (IH, m), 4-50 (lH, m). (Found: C, 72.71; H, 10.30. Calc. for Cr,H,p,: C, 72.82; H, 10.06%). (iii) Further elution with CHCI, (300 ml) gave 19 (0.4 g) as needles, m.p. and mixed m.p. 316-318O (open caphary). Acetylation of 19 (100 mg) with Ac,O (1 ml) and pyridine (2 ml) gave 36. 156, 2la-rrloceroxy-14b hydroaysewatane (21; 95 Ing) as needles from benzene-n-hexane, m-p. 297-299’. (Found: C, 71.68; H, 9.63. Calc. for Ca,,O,: C, 71.72; H, 9.70%). Acetylation of 20 (100 mg) with AcsO (1 ml) and pyridine (2 ml) gave 22 (97 me) as needles from bcnxene-n-hexanc, m.p. 257-259“. (Found: C, 71.64; H, 9.77. Clac. for C,$I&,: C, 71.72; H, 9.70%). Serini reaction of 16
Compound 16 (12 mg) and Zn dust (242 mg) were mixed and sublimed at 250’ and 1 mmHg. The TLV and IR of the crystalline sublimate showed that the product was a mixture of Ma and 14@-serratanM-ones, the major product being 6
Stereochemistry of addition reaction to serratene and serratenediol
759
Serini reaction of 21 and 22 (i) Compound 21 (60 mg) and Zn dust (1 g) in toluene (10 ml) were heated under retlux for 15 hr. removal of Zn dust and solvent letI a crystalline residue which was crystallized from CHCls-MeOH to give 38,2ladiac&q-l4a-~watcmtl5-one (23; 30 mg) as prisms, m.p. 296296’ (open capillary); IR 1725 @AC) and 1960 cm-’ (ketone); NMR (6): 0.85 (12H, s), 0.90 (3H. s). 0.92 (3H, s), 1.06 (3H, s), 2.05 (6iH, s). 4.45 (W, m). (Found: C, 75.52; H. 10.13. Calc. for C&,0,: C, 75.23; H, 10.03%). (ii) Compound 22 (60 mg) and Zn dust (1 g) in toluene (10 ml) were similarly treated Crystallization of the product from CHCI,-MeCH afforded 23 (20 mg) m.p. and mixed m.p. 296296”. (iii) A mixture of 21(10 mg) and Zn dust (120 mg) was sublimed at 270° and 1 mm. The TLC of the crystalline sublimate showed a single spot corresponding to 23. The IR spectrum was superimposable on that of the sample obtahred above. (iv)Amixhrreof22(11mg)andZndust(17lmg)wassublimedasdescribedabove.The~Cofthe sublimate showed two spots, one of which was identical with that of 23. The IR spwtntm of the mixture exhibited a CO band at 1685 cm-*, but no OH band. Isomcrixutfon of 14$-serratmr-l5-one under the condition #the Serid reaction Compound 5 (10 mg) and Zn dust (500 mg) in toluene (10 ml) containing 2 drops of AcOH. was re5uxed for 5 hr. The TLC of the product showed two spots corresponding to 6 and 5. 38,2 la-Diuceto~-l4a-smatan-l5-one (23) (i)Asolnof19(17mg)in3%HCCEtOH(lOml)washeatcdundarcfluxforlhr.Themixturcwas neutralized with 5% KxCO,aq and extracted with CHCI,. The extract was washed and water, dried over MgSO, and evaporated to dryness. The residue was acetylated with AcrO (1 ml) and pyridine (2 ml) at room temp to give 23 (11 mg), m.p. and mixed m.p. 294-296’ (open capillary). (ii) A sohr of 22 (40 mg) in 3 % HCI-EtOH (20 ml) and CHCI, (2 ml) was treated as described above. Acetylation of the product gave 23 (26 mg). m.p. 294-296’ (open capillary). 38, 21~Dhcetary-14B, 15~qrserrctune (18) (i) Coupound 17’ (2.4 g) in CHCI, (30 ml) was treated with an ethereal sohr of monoperphthahc acid (large excess) at room temp for a week. Tbe reaction mixture was dilti with CHCI, and the organic layer was washed with 5% NaOH aq, water, dried over MgSO, and evaporated to dryness. Crystallixation of the residue from benzene afforded the oxide (18) as prisms (1.58 g). m.p. 350-353” (open capifbuy), [a]n +27.8” (c = 36 in CHCI,); NMR (a): 079 (3H s), 090 (12H s), 097 (3H, s), 1W (3H, s), 2.10 (6H, s), 2.88 (broad s), 450 (ZH, m) (Found: C, 7508; H, 9.83. Calc for Cs,H,,O, : C, 75.23; H, 10.03%). (ii) Oxonixed 0, was passed through a soln of 17 (728 mg) in CHCI, (15 ml) at 0’ until the mixture was s&mated to tetranitromethane. The mixture was diluted with water, extramd with CHCI, and the extract was washed with water, dried over MgSO, and evaporated to dryness. Crystallization of the residue from CHCI,-MeCH gave 18 (600 mg) as prisms, m.p. and mixed m.p. 340-343”. 15-oxoneoserrataIu? (26) Compound 3 (60 mg) in CHCl, (I5 ml) was treated with a few drops of cone HCl under stirring for 30 min. and the product in n-hexane was chromatographed over alumina (1 g). The Erst ehmte afforded a hydrocarbon mixture (10 mg), 1, (EtCH): 23 1,241,249 and 280 mp, further elution with n-hexane and benzene gave the aldehyde (W, 35 mg) as prisms from acetone, m.p. 17617S”; IR: 2703 and 1718 cm-’ (CHO); NMR (a): 0.70 (3H, s), 0.76 (9H. s), 0.79 (12H. s), 9.50 (lH, s). (Found: C, 84.68; H, 11.98. Calc. for C,&,O: C, 8444; H, 11.81%). 38,2la-Diacet~l5-oxoneoserrctane (27) (i)Compounds 18(198mg)inCHC1,(15ml)wastreatedwithconcHC1(0~1mf)atroomtempfor30 min with stirring. Crystallization of the product kom benxene atTorded the uldehyak (27; 90 mg) as needles, m.p. 245-247”; [a], +21.6” (c = 3.43 in CHClJ; IR: 1730 cm-* (CHO, OAc); NMR (a): 0.75 (3H, s), 0.82 (3H, s), 0.88 (9H, s), 0.96 (6H. s), 2.09 (6H. s), 450 (W. m). 9.55 (lH, s). (Found: C, 75.14: H, 9.84. Calc. for CjI,,O,: C, 75.23; H, 10.03%). (ii)Toasolnof17(500mg)inAcOEt(5Oml)wasaddcdasdnof8O%HCOOH(l0ml)and3O% H,O,42ml),andthemixtunwasheatedundcrreeuxfor lhr,and concenkatedtoca.+vohuneinvatxW.
760
Y. TSUDA,T. SANOand Y.
hUBUSHI
Collection of the ppt and crystallization from CHCI,MeOH gave 27 (260 mg), as leaflets, m.p. and mixed m.p. 24s246O. (iii) Compound 17 (728 mg) in CHCI, (15 ml) was ozonized as described above. Zn dust (1.1 g) and AcOH (8 ml) were added to the mixture, and after. stirring for 2 hr the ppt was Eltered off and washed with CHCI, The combined filtrate and washings were washed with water, dried over MgSO, and evaporated to dryness. Chromatography of the product and crystallization from CHCl,MeOH gave 27 (350 mg), m.p. and mixed m.p. 247-248”. Hydrolysis of 27 (100 mg) with 2% KOH-MeOH (30 ml) under refluz for 1 hr afforded the corresponding diol (70 mg) as needles from MeOH m.p. 209-211°; IR: 3390 cm-* (OH), 1721 (CHO). (Found : C, 7550; H, 10.89. Calc. for C,,,H,,O, . HsO : C, 75.58 ; H, 11GCl %). 3 B, 2 la-Diaceto~eoserratan- 15-oic acid (33) To a soln of 27 (290 mg) in pyridine (10 ml) was added KMnO, (240 mg) and the mixture was stirred for 7 hr at room temp. After decomposition of KMnO, by MeOH, the solvent was evaporated to dryness in D(UIUO. The residue was taken up in CHQ, and the extract was washed with Na&O, and dil HsSO,, water, dried over MgSO,, and evaporated to dryness Crystallization of the residue from MeOH afforded the crcid (33; 170 mg) as needles, m.p. 285-2XP, [o]u -5.1” (c = 2.7 in CHCl,); IR: 32541705 (COOH), and 1726 cm-’ (OAc). (Found: C, 72.95; H, 9.85. Calc. for CJi,,O,: C, 73.08; H, 9.74%). Tbe acid 33 (110 mg) was methylated with excess of CHsNN,in ether. Crystallization of the product from CHCl,-MeOH gave the methyl ester (34; 85 mg) as needles, m.p. 232-234’, [a], -17. lo (c = 2.8 in CHCls); IR: 1730 cm-’ (COOMe, OAc). (Found: C, 7328; H, 1011. Calc for CssH,eOs: C, 73.39; H, 9.85%). 3 b 2 la-Dihydrogmeoserratane-15-o& acid (35) The diacetate 33 (115 mg) was hydrolysed with 10% KOH-MeOH (15 ml) under reiluz for 4 hr. The mixture was poured into water, and crystallization of the pp1 from CHCl,MeOH gave the dti (36; 85 mg) as needles, m.p. 345-347O (open capillary). (Found: C, 73.42; H, 10.72. Calc. for CjI,O,.H,O: C, 73.12; H, 10.64%). The acid 35 (320 mg) in CHCI,-MeOH (1:l) (30 ml) was methylated with CHsNN,in ether. Crystallization of the product from CHCI,-MeOH gave the methyl ester (36; 292 mg) as needles, m.p. 255257’; IR: 3390, 3322 (OH) and 1718 cm-’ (eater). (Found: C, 74.49; H, 10.83. Calc, for CJ-&,O,.+ H,O: C, 74.79; H, 10.73%). The diketoesfer (37) A mixture of 36 (200 mg) and 00, (400 mg) in pyridine (5 m,) was kept overnight at room temp. The product, isolated with CHCI, was chromatographed in benzene over alumina (2 g), and the eluate was crystallized from n-hexane to give the diketo-ester(37) as needles, m.p. 203-205O, [al, +35.6” (c = 3.6 in CHCI,); IR: 1709 cm-’ (ester and ketone). (Found: C, 7658; H, 991. Caic for Cs,H,O4: C 7681; H, 9.98 %). 3 g, 2 la-Dihydroxyneoserratwuz(28) Compound 27 (500 mg) and anhyd hydraxine (4 ml) in diethykneglycol(5O ml) containing Na (4 g) were heated at 180” for 4 hr, then the temp was raised to 210” and the mixture was kept for 4 hr. The mixture was poured into water, extracted with CHCl, and the extract was washed with water, dried over MgSO, and evaporated to dryness. Crystallixation of the residue (420 mg) from CHCI,-MeOH gave 28 as prisms, m.p. 265-267O. (Found: C. 79.07; H, 11.52. Calc. for CJI,,O,+H,O: C, 79.40; H, 11.78%). Acetylation of this (100 mg) with AC,0 (1 ml) and pyridirc (2 ml) gave the diucetute (92 mg) as needles from CHCl,MeOH, m.p. 256257”. (Found: C, 77.40; H, 10.88. Calc. for CJi,p,: C, 77.22; H, 10.67%). Neoserrotaw (30)
Compound 28 (200 mg) and 00s (300 mg) in pyridine (5 ml) were stirred overnight at room temp. The mixture was poured into water and extracted with C!HC&.The extract was washed with water, dried over Mg!SO*,and evaporated to dryrxsa The residue was chromatographal in CHCls over silica-gel to
Stereochemistry of addition reaction to serratetie and serratenediol
761
give neoserrarun-3,2l-&ne (299; 160 mg) (needles from CHCl,-MeOH), m.p. 140-143”. IR: 1695 cm-’ cyclobexanone). Compound 29 (80 mg) and anhyd hydrazinc (1 ml) in diethyleneglycol(20 ml) containing Na (1 g) were beated as described above. Chromatography of the product in n-hexane over A1203 gave 30 (40 mg) (leaflus from CHCl,-M&H). (Found: C, 87.08; H. 12.56. Calc. for C&I,,: C, 87.30; H, 12.70%). Methyl neawrratan- 1S-oate (32) (i) The diketo-&er 37 (623 mg) and anhyd hydraxine (6 ml) in diethyleneglycol(42 ml) containing Na (3 g) were heated at MO0 for 4 hr, and heating was continued at 210” for further 2 hr. The mixture was poured into water and extracted with CHCI, The extract was washed with water, dried over MgSO, and evaporated to dryness. The residue in ether (10 ml) was treated with CHxNN,to give the methyl ester (32; 510 mg) as needles from MeOH, m.p. 149-151°; IR: 3509 (OH) and 1724 cm-’ (COOMe). (Found: C, 79.89; H, 11.49. Calc. for C,,H,xO,.+MeOH: 80.03; H, 11.50%). (ii) The aldehyde 26 (100 mg) was oxidized with KMnO, in pyridine as described for 33. Methylation of the crude product with CHgN, afforded the methyl ester 32, m.p. and mixed m.p. 149-151°. Oxidation of serratene (1) with chromium trioxiak Smatene 1(467 mg) in AcOH (150 ml) and 00, (500 mg) in 90% AcOH (10 ml) were mixed and the mixture was warmed at 75O on a water bath for 3 hr. At&r decomposition of excess of 00, with EtOH, AcOH was removed &I vucno, and the residue was taken up in CHCI, which was washed with water, dried over MgSO, and evaporated to dryness. Washing of the residue with n-hexane gave a crystalline residue (225 mg). Chromatography of this and crystallixation of benzene and CHCI, eluates gave neosewatnn-15-o& acti (31; 210 mg) (Rims from CHCl,-MeOH), m.p. >300°; IR: 1690 cm-’ (COOH). (Found: C, 81.42; H, 11.34. Calc. for CjI,O,: C, 81.39; H, 11.38%). Treatment of 31(50 mg) with CHflN, in ether gave 32 (42 mg) as needles from CHCI,-MeOH, m.p. and mixed m.p. 149-151’. The hexane-soluble oil (285 mg) was passed through a column of alumina and the eluate was crystallized from acetone to give serrat-13-en-15one’ as leaflets, m.p. and mixed m.p. 245-246’. REFERENCES i Y. Inubushi, T. Tsuda, T. Sane, T. Konita, S. Suzuki, H. Ageta and Y. Otake, Chem. Phunn. Bull. Tdyo 15, 1153 (1967). * A. S. Hallsworth and H. B. Henbest, 1. Chem. Sot. 4604 (1957). a M. Hashimoto and Y. Tsuda, Reiiminmy Report of NMR Symposium, M-2-13, Tokyo, September 1 (1965). ’ Organic Reactions Vol. 13, p 1, Wiley, N.Y. (1963). ’ L. F. Fieser and M. Fieser. S:ervids p 628. Reinhold N.Y. (1959). 6 J. W. ApSimon, P. V. Demarco and I. Len&e, Camad. J. Chin, 43.2793 (1965). ’ Y. Tsuda, K. Isobe, S. Fukushima, H. Ageta and K. Iwata, Terrohedron Letters 23 (1967). * J. M. Beaton, F. S. Spring, R. Stevenson and J. L. Stewart,I. Chem. Sot. 2131 (1955).